一、IPv6编址
1、基础
IPv6地址有128位，将这128位分成8组，每组16位，中间用:隔开；第一个地址为全0，最后一个地址为全12
每一组都用16进制表示，范围是0000-----FFFF
2、简写规则
3、IPv6地址空间分类
| 整个IPv6地址空间 | ||
| 128位中的第一组(高16位) | 16进制表示 | 备注 |
| 0000 0000 xxxx xxxx | 0000-00FF | 保留、环回、兼容地址 |
| 0000 0001 xxxx xxxx | 0100-01FF | 未定义用途 |
| 0000 001x xxxx xxxx | 0200-03FF | 未定义用途 |
| 0000 01xx xxxx xxxx | 0400-07FF | 未定义用途 |
| 0000 1xxx xxxx xxxx | 0800-0FFF | 未定义用途 |
| 0001 xxxx xxxx xxxx | 1000-1FFF | 未定义用途 |
| 001x xxxx xxxx xxxx | 2000-3FFF | 全球可聚合单播地址 |
| 01xx xxxx xxxx xxxx | 4000-7FFF | 未定义用途 |
| 10xx xxxx xxxx xxxx | 8000-BFFF | 未定义用途 |
| 110x xxxx xxxx xxxx | C000-DFFF | 未定义用途 |
| 1110 xxxx xxxx xxxx | E000-EFFF | 未定义用途 |
| 1111 0xxx xxxx xxxx | F000-F7FF | 未定义用途 |
| 1111 10xx xxxx xxxx | F800-FBFF | 未定义用途 |
| 1111 110x xxxx xxxx | FC00-FDFF | 唯一本地地址(私网地址) |
| 1111 1110 0xxx xxxx | FE00-FE7F | 未定义用途 |
| 1111 1110 10xx xxxx | FE80-FEBF | 链路本地地址 |
| 1111 1110 11xx xxxx | FEC0-FEFF | 站点本地地址(已废止) |
| 1111 1111 xxxx xxxx | FF00-FFFF | 组播 |
最重要三类地址:全球可聚合单播地址(类似IPv4中的公网地址)、链路本地地址以及组播
4、单播地址构成
链路本地地址(Link-local address):用于自动地址配置、邻居发现、路由器发现等。
节点启用IPv6后，会自动生成Link-local address地址: FE80:0:0:0:接口ID或FE80::接口ID
5、组播地址构成
6、组播MAC地址构成
7、常用组播地址
| IPv4定义组播地址 | IPv6定义组播地址 | 备注 | 组播mac地址 |
| 节点本地范围 | |||
| 224.0.0.1 | FF01::1 | 所有节点 | 33:33:0:0:0:1 |
| 224.0.0.2 | FF01::2 | 所有路由器 | 33:33:0:0:0:2 |
| 链路本地范围 | |||
| 224.0.0.1 | FF02::1 | 所有节点 | 33:33:0:0:0:1 |
| 224.0.0.2 | FF02::2 | 所有路由器 | 33:33:0:0:0:2 |
| 224.0.0.5 | FF02::5 | 所有OSPF路由器 | 33:33:0:0:0:5 |
| 224.0.0.6 | FF02::6 | 所有OSPF DR路由器 | 33:33:0:0:0:6 |
| 224.0.0.9 | FF02::9 | 所有RIP路由器 | 33:33:0:0:0:9 |
| 224.0.0.13 | FF02::D | 所有PIM路由器 | 33:33:0:0:0:D |
| 站点本地范围 | |||
| 224.0.0.2 | FF05::2 | 所有路由器 | 33:33:0:0:0:2 |
8、特殊组播地址-被请求节点组播地址(主要用于重复地址检测（DAD）和替代IPv4中的ARP)
每一个单播地址都有相对应的被请求节点组播地址。被请求节点组播地址前104为被固定，后24位为单播地址的后24位。被请求节点组播MAC地址前24位固定33:33:FF，24位为单播地址的后24位。
二、IPv6节点
三、ICMPv6协议
四、IPv6地址配置方法
五、NDP邻居发现协议
1、地址解析-通过邻居请求（NS）和邻居通告（NA）报文来解析三层地址对应的链路层地址。
(IPv6中没有广播，没有ARP)
B、PC_A与PC_B正常通信
C、PC_A与PC_C、PC_B与PC_C之间通信过程与上一致
2、无状态自动配置-地址前缀公告
PC_A、PC_B、PC_C获取路由前缀途径1: PC_A、PC_B、PC_C分别发送RS报文，，Router回应RA报文
PC_A、PC_B、PC_C获取路由前缀途径2: Router周期性发送RA报文，无需接收到RS报文
路由器通告路由前缀的前提：
1、接口启用IPv6
2、接口配置了唯一本地地址或者全球可聚合单播地址
路由器通告路由前缀信息：
1、路由前缀长度默认是64位
2、路由前缀的有效期限
preferred lifetime:能主动访问，也能被动接受访问
Valid lifetime： 在preferred lifetime之外，Valid lifetime有效期内只能被动接受访问
3、提供给节点的网关为路由器接口的本地链路地址，非接口唯一本地地址或者全球可聚合单播地址
2、无状态自动配置-地址重复检测DAD
节点启动时：
A、PC_A、PC_B和PC_C节点启动后，自动生成链路本地地址
PC_A:FE80::0323:45FF:FE67:89AA
PC_B:FE80::0323:45FF:FE67:89BB
PC_C:FE80::0323:45FF:FE67:89CC
B、在完成对链路本地地址DAD检测前，链路本地地址不能进行通信
C、DAD检测(三台主机分别发送NS报文)
PC_A发送NS报文，请求FE80::0323:45FF:FE67:89AA的MAC地址，如果收到NA报文，表明该地址已经被使用了；没有收到任何回应，则表明PC_A可以使用FE80::0323:45FF:FE67:89AA地址，DAD检测通过
D、三台主机分别发送NS报文，均没有收到NA报文后，分别发送NA组播报文，宣告自己的IP地址(组播地址为FF02:1----本链路范围所有节点)
节点获取前缀后：
A、PC_A、PC_B和PC_C节点获取前缀2008::/64后，构造可聚合单播地址
B、在完成对可聚合单播地址DAD检测前，可聚合单播地址不能进行通信
C、DAD检测(三台主机分别发送NS报文)
PC_A发送NS报文，请求2008::0323:45FF:FE67:89AA的MAC地址 ,如果收到NA报文，表明该地址已经被使用了；没有收到任何回应，则表明PC_A可以使用2008::0323:45FF:FE67:89AA地址，DAD检测通过
D、三台主机分别发送NS报文，均没有收到NA报文后，分别发送NA组播报文，宣告自己的IP地址(组播地址为FF02:1----本链路范围所有节点)
2、无状态自动配置-前缀重新编址
3、路由重定向
A、PC_A上到2012::FFFF/64的路由，设置RouterB为下一跳
B、 RouterB上到2012::FFFF/64的路由，设置RouterA为下一跳
C、 RouterB发送一个ICMPv6 Type=137的重定向报文给PC_A,
去往2012::FFFF/64，应该将数据包发送给RouterA
D、 PC_A收到RouterB发来的重定向报文后，设置RouterA为到目的地址2012::FFFF/64的下一跳
4、邻居状态跟踪
六、IPv4到IPv6过渡技术
七、实验
实验一、无状态自动配置(地址前缀公告、地址重复检测DAD)
RouterA: ipv6 nd ra suppress //路由器不主动发送RA,只有接收到RS报文时才发送RA
可聚合单播地址：2012::FFFF/64
e0/0接口MAC:ca04.2cd5.0006
linklocal地址：FE80::C804:2CFF:FED5:6
组播地址：FF021 FF02:2
被请求节点组播：FF021:FF00:FFFF
被请求节点组播：FF02::1:FFD5:6
组播MAC：3333:FF00:FFFF /3333:FFD5:0006
RouterB: ipv6 nd ra suppress //路由器不主动发送RA,只有接收到RS报文时才发送RA
可聚合单播地址：2112::FFFF/64
e0/0接口MAC:ca03.2e25.0006
linklocal地址：FE80:C803:2EFF:FE25:6
组播地址：FF02:1 FF02:2
被请求节点组播：FF02:1:FF00:FFFF
被请求节点组播：FF021:FF25:6
组播MAC：3333:FF00:FFFF 3333:FF25:0006
Client1: ipv6 address autoconfig default
e0/0接口MAC:ca01.7434.0006
linklocal地址:FE80:C801:74FF:FE34:6
组播地址：FF021
被请求节点组播：FF02::1:FF34:6
组播MAC:3333-FF34-0006
Client2: ipv6 address autoconfig default
e0/0接口MAC:ca05.7564.0006
linklocal地址:FE80:C805:75FF:FE64:6
组播地址：FF02:1
被请求节点组播：FF02::1:FF64:6
组播MAC:3333-FF64-0006
client1和client2获取路由器前缀过程：(以client1为例)
1、client1基本信息
e0/0接口MAC:ca01.7434.0006
linklocal地址:FE80:C801:74FF:FE34:6
组播地址：FF021
被请求节点组播：FF02::1:FF34:6
组播MAC:3333-FF34-0006
2、DAD检测-针对linklocal地址
client1发送NS报文请求linklocal地址FE80:C801:74FF:FE34:6的MAC地址
3、如果链路上有节点使用该地址，会发送NA报文回应；client1没有收到NA报文，然后发送组播报文NA，宣告自己使用该地址FE80:C801:74FF:FE34:6,DAD检测完成
4、client1发送RS报文请求路由器前缀
目的地址为FF02::2,本链路范围所有路由器组播地址
5、RouterA收到RS报文后发送RA单播报文回应client1
路由器A分配给的前缀是2012::/64，同时还有两个lifetime-valid和preferred
6、client1收到RouterA的RA报文后，做DAD检测(发送NS报文)
用路由器A给的前缀构造IPv6地址2012:C801:74FF:FE34:6后，也要对该地址做DAD检测.
client1没有收到NA报文，然后发送组播报文NA，宣告自己使用该地址2012::C801:74FF:FE34:6,DAD检测完成。
7、client1发送RS报文请求路由器前缀
目的地址为FF02::2,本链路范围所有路由器组播地址，除了RouterA收到RS请求然后给cilent1发送RA报文外，RouterB也收到client1发出的RS请求报文
8、RouterB发送RA单播报文回应client1
9、client1收到RouterB的RA报文后，做DAD检测(发送NS报文)
10、client1发送NA报文，通告自己使用的单播地址2112::C801:74FF:FE34:6
11、完成DAD检测后，可用单播地址通信
client1获取地址如下:
路由如下：
client2获取地址如下:
路由如下：
实验二、无状态自动配置(前缀重新编址)
修改RouterB的前缀Valid lifetime和preferred lifetime:
int e0/0
ipv6 add 2222::/64
ipv6 nd prefix 2112::/64 30 15
ipv6 nd prefix 2222::/64 43200 43200
ipv6 nd ra suppress
修改前client1：do show ipv int e0/0
修改后client1：do show ipv int e0/0
2112前缀地址已经不能发起链接了，只能被动接收请求；等Valid lifetime计时为0，该前缀将彻底失效。
实验三、地址解析
client1与client2通信，需要知道client2的MAC地址
client1只有RouterA和RouterB的MAC
现在从client1上ping client2:
ping FE80::C805:75FF:FE64:6
client1先发送NS组播报文 ，请求FE80::C805:75FF:FE64:6的MAC地址
client1接收NA报文
查看邻居状态
实验四、邻居状态跟踪
debug ipv6 icmp
debug ipv6 nd
Client1#show ipv neighbors
IPv6 Address Age Link-layer Addr State Interface
FE80::C803:2EFF:FE25:6 130 ca03.2e25.0006 STALE Et0/0
FE80::C804:2CFF:FED5:6 130 ca04.2cd5.0006 STALE Et0/0
Client1#ping 2012::FFFE repeat 1
Type escape sequence to abort.
Sending 1, 100-byte ICMP Echos to 2012::FFFE, timeout is 2 seconds:
!
Success rate is 100 percent (1/1), round-trip min/avg/max = 88/88/88 ms
Client1#
*Mar 25 21:25:07.361: ICMPv6: Sent echo request, Src=2012::C801:74FF:FE34:6, Dst=2012::FFFE
*Mar 25 21:25:07.361: ICMPv6-ND: DELETE -> INCMP: 2012::FFFE
*Mar 25 21:25:07.365: ICMPv6-ND: Sending NS for 2012::FFFE on Ethernet0/0
*Mar 25 21:25:07.365: ICMPv6-ND: Resolving next hop 2012::FFFE on interface Ethernet0/0
*Mar 25 21:25:07.369: ICMPv6: Sent N-Solicit, Src=2012::C801:74FF:FE34:6, Dst=FF02::1:FF00:FFFE
*Mar 25 21:25:07.413: ICMPv6: Received N-Advert, Src=2012::FFFE, Dst=2012::C801:74FF:FE34:6
*Mar 25 21:25:07.413: ICMPv6-ND: Received NA for 2012::FFFE on Ethernet0/0 from 2012::FFFE
*Mar 25 21:25:07.417: ICMPv6-ND: Neighbour 2012::FFFE on Ethernet0/0 : LLA ca03.2e25.0006
*Mar 25 21:25:07.417: ICMPv6-ND: INCMP -> REACH: 2012::FFFE
*Mar 25 21:25:07.453: ICMPv6: Received echo reply, Src=2012::FFFE, Dst=2012::C801:74FF:FE34:6
*Mar 25 21:25:12.493: ICMPv6-ND: STALE -> DELAY: FE80::C803:2EFF:FE25:6
*Mar 25 21:25:17.537: ICMPv6-ND: DELAY -> PROBE: FE80::C803:2EFF:FE25:6
*Mar 25 21:25:17.537: ICMPv6-ND: Sending NS for FE80::C803:2EFF:FE25:6 on Ethernet0/0
*Mar 25 21:25:17.541: ICMPv6: Sent N-Solicit, Src=FE80::C801:74FF:FE34:6, Dst=FE80::C803:2EFF:FE25:6
*Mar 25 21:25:17.565: ICMPv6: Received N-Advert, Src=FE80::C803:2EFF:FE25:6, Dst=FE80::C801:74FF:FE34:6
*Mar 25 21:25:17.565: ICMPv6-ND: Received NA for FE80::C803:2EFF:FE25:6 on Ethernet0/0 from FE80::C803:2EFF:FE25:6
*Mar 25 21:25:17.565: ICMPv6-ND: PROBE -> REACH: FE80::C803:2EFF:FE25:6
Client1#show ipv neighbors
IPv6 Address Age Link-layer Addr State Interface
FE80::C803:2EFF:FE25:6 0 ca03.2e25.0006 REACH Et0/0
FE80::C804:2CFF:FED5:6 131 ca04.2cd5.0006 STALE Et0/0
2012::FFFE 0 ca03.2e25.0006 REACH Et0/0
*Mar 25 21:25:37.517: ICMPv6-ND: REACH -> STALE: 2012::FFFE
*Mar 25 21:25:47.629: ICMPv6-ND: REACH -> STALE: FE80::C803:2EFF:FE25:6
Client1#show ipv neighbors
IPv6 Address Age Link-layer Addr State Interface
FE80::C803:2EFF:FE25:6 0 ca03.2e25.0006 STALE Et0/0
FE80::C804:2CFF:FED5:6 131 ca04.2cd5.0006 STALE Et0/0
2012::FFFE 0 ca03.2e25.0006 STALE Et0/0
Client1#ping 2012::FFFE repeat 1
Type escape sequence to abort.
Sending 1, 100-byte ICMP Echos to 2012::FFFE, timeout is 2 seconds:
!
Success rate is 100 percent (1/1), round-trip min/avg/max = 8/8/8 ms
Client1#
*Mar 25 21:26:02.197: ICMPv6: Sent echo request, Src=2012::C801:74FF:FE34:6, Dst=2012::FFFE
*Mar 25 21:26:02.201: ICMPv6-ND: STALE -> DELAY: 2012::FFFE
*Mar 25 21:26:02.209: ICMPv6: Received echo reply, Src=2012::FFFE, Dst=2012::C801:74FF:FE34:6
*Mar 25 21:26:02.213: ICMPv6-ND: ULP indication 2012::FFFE/Ethernet0/0
*Mar 25 21:26:02.213: ICMPv6-ND: DELAY -> REACH: 2012::FFFE
*Mar 25 21:26:07.245: ICMPv6: Received N-Solicit, Src=FE80::C803:2EFF:FE25:6, Dst=2012::C801:74FF:FE34:6
*Mar 25 21:26:07.249: ICMPv6-ND: Received NS for 2012::C801:74FF:FE34:6 on Ethernet0/0 from FE80::C803:2EFF:FE25:6
*Mar 25 21:26:07.249: ICMPv6-ND: Sending NA for 2012::C801:74FF:FE34:6 on Ethernet0/0
*Mar 25 21:26:07.253: ICMPv6: Sent N-Advert, Src=2012::C801:74FF:FE34:6, Dst=FE80::C803:2EFF:FE25:6
*Mar 25 21:26:07.253: ICMPv6-ND: STALE -> DELAY: FE80::C803:2EFF:FE25:6
*Mar 25 21:26:12.325: ICMPv6-ND: DELAY -> PROBE: FE80::C803:2EFF:FE25:6
*Mar 25 21:26:12.325: ICMPv6-ND: Sending NS for FE80::C803:2EFF:FE25:6 on Ethernet0/0
*Mar 25 21:26:12.325: ICMPv6: Sent N-Solicit, Src=FE80::C801:74FF:FE34:6, Dst=FE80::C803:2EFF:FE25:6
*Mar 25 21:26:12.333: ICMPv6: Received N-Advert, Src=FE80::C803:2EFF:FE25:6, Dst=FE80::C801:74FF:FE34:6
*Mar 25 21:26:12.337: ICMPv6-ND: Received NA for FE80::C803:2EFF:FE25:6 on Ethernet0/0 from FE80::C803:2EFF:FE25:6
*Mar 25 21:26:12.341: ICMPv6-ND: PROBE -> REACH: FE80::C803:2EFF:FE25:6
*Mar 25 21:26:17.397: ICMPv6: Received N-Solicit, Src=FE80::C803:2EFF:FE25:6, Dst=FE80::C801:74FF:FE34:6
*Mar 25 21:26:17.397: ICMPv6-ND: Received NS for FE80::C801:74FF:FE34:6 on Ethernet0/0 from FE80::C803:2EFF:FE25:6
*Mar 25 21:26:17.397: ICMPv6-ND: Sending NA for FE80::C801:74FF:FE34:6 on Ethernet0/0
*Mar 25 21:26:17.401: ICMPv6: Sent N-Advert, Src=FE80::C801:74FF:FE34:6, Dst=FE80::C803:2EFF:FE25:6
Client1#show ipv neighbors
IPv6 Address Age Link-layer Addr State Interface
FE80::C803:2EFF:FE25:6 0 ca03.2e25.0006 REACH Et0/0
FE80::C804:2CFF:FED5:6 132 ca04.2cd5.0006 STALE Et0/0
2012::FFFE 0 ca03.2e25.0006 REACH Et0/0
*Mar 25 21:26:32.225: ICMPv6-ND: REACH -> STALE: 2012::FFFE
*Mar 25 21:26:42.417: ICMPv6-ND: REACH -> STALE: FE80::C803:2EFF:FE25:6
Client1#show ipv neighbors
IPv6 Address Age Link-layer Addr State Interface
FE80::C803:2EFF:FE25:6 0 ca03.2e25.0006 STALE Et0/0
FE80::C804:2CFF:FED5:6 132 ca04.2cd5.0006 STALE Et0/0
2012::FFFE 0 ca03.2e25.0006 STALE Et0/0
Client1#ping 2012::FFFE repeat 1
Type escape sequence to abort.
Sending 1, 100-byte ICMP Echos to 2012::FFFE, timeout is 2 seconds:
*Mar 25 21:26:54.817: ICMPv6: Sent echo request, Src=2012::C801:74FF:FE34:6, Dst=2012::FFFE
*Mar 25 21:26:54.817: ICMPv6-ND: STALE -> DELAY: 2012::FFFE.
Success rate is 0 percent (0/1)
*Mar 25 21:26:59.893: ICMPv6-ND: DELAY -> PROBE: 2012::FFFE
*Mar 25 21:26:59.893: ICMPv6-ND: Sending NS for 2012::FFFE on Ethernet0/0
*Mar 25 21:26:59.897: ICMPv6: Sent N-Solicit, Src=FE80::C801:74FF:FE34:6, Dst=2012::FFFE
*Mar 25 21:27:00.909: ICMPv6: Received R-Advert, Src=FE80::C804:2CFF:FED5:6, Dst=FF02::1
*Mar 25 21:27:00.989: ICMPv6-ND: Sending NS for 2012::FFFE on Ethernet0/0
*Mar 25 21:27:00.993: ICMPv6: Sent N-Solicit, Src=FE80::C801:74FF:FE34:6, Dst=2012::FFFE
*Mar 25 21:27:02.081: ICMPv6-ND: Sending NS for 2012::FFFE on Ethernet0/0
*Mar 25 21:27:02.081: ICMPv6: Sent N-Solicit, Src=FE80::C801:74FF:FE34:6, Dst=2012::FFFE
*Mar 25 21:27:03.173: ICMPv6-ND: PROBE deleted: 2012::FFFE
*Mar 25 21:27:03.173: ICMPv6-ND: PROBE -> DELETE: 2012::FFFE
Client1#show ipv neighbors
IPv6 Address Age Link-layer Addr State Interface
FE80::C803:2EFF:FE25:6 1 ca03.2e25.0006 STALE Et0/0
FE80::C804:2CFF:FED5:6 133 ca04.2cd5.0006 STALE Et0/0
Client1#undebug all
All possible debugging has been turned off
实验五、路由重定向
RouterA:可以主动发送RA
interface Ethernet0/0
no ip address
duplex auto
ipv6 address 2012::FFFF/64
ipv6 enable
ipv6 nd ra interval 10
RouterB:不主动发送RA，也不进行前缀通告
interface Ethernet0/0
no ip address
duplex auto
ipv6 address 2012::FFFE/64
ipv6 enable
ipv6 nd prefix 2012::/64 no-advertise
ipv6 nd ra suppress
client1: ipv6 route 3000::/64 e0/0 FE80::C803:2EFF:FE25:6
RouterB: ipv6 route 3000::/64 e0/0 FE80::C804:2CFF:FED5:6
RouterA:
int loopback0
ipv6 enable
ipv6 add 3000::FFFF/64
Client1#ping ipv6 3000::FFFF source 2012::C801:74FF:FE34:6
RouterB发送重定向报文给client1
同时RouterB将原始ping报文转发给RouterA，RouterA收到 ping报文后，RouterA发送NS请求2012::C801:74FF:FE34:6的mac地址，client1发送NA给RouterA告诉其2012::C801:74FF:FE34:6的mac地址，最后RouterA发送回包给client1
后续Client1#ping ipv6 3000::FFFF source 2012::C801:74FF:FE34:6包，不会再发给RouterB了
即使Client1上有路由ipv6 route 3000::/64 Ethernet0/0 FE80::C803:2EFF:FE25:6指向RouterB
实验六、隧道技术
1、手工隧道------IPv6 over IPv4隧道
interface Tunnel0
no ip address
ipv6 enable
tunnel source e0/1
tunnel mode ipv6ip
tunnel destination 10.10.23.3
ip route 0.0.0.0 0.0.0.0 10.10.12.2
ipv6 route ::/0 Tunnel0
interface Tunnel0
no ip address
ipv6 enable
tunnel source e0/2
tunnel mode ipv6ip
tunnel destination 10.10.12.1
ip route 10.10.12.0 255.255.255.0 10.10.23.2
ipv6 route ::/0 Tunnel0
IPv6整个数据包作为载荷被封装在IPv4包中
2、手工隧道------GRE隧道
interface Tunnel0
no ip address
ipv6 enable
tunnel source e0/1
tunnel mode gre ip
tunnel destination 10.10.23.3
ip route 0.0.0.0 0.0.0.0 10.10.12.2
ipv6 route ::/0 Tunnel0
interface Tunnel0
no ip address
ipv6 enable
tunnel source e0/2
tunnel mode gre ip
tunnel destination 10.10.12.1
ip route 10.10.12.0 255.255.255.0 10.10.23.2
ipv6 route ::/0 Tunnel0
IPv6包封装在GRE中，GRE包封装在IPv4包中
3、自动隧道-----6to4隧道
IPv6固定格式：2002:公网IPv4:子网::/64
注意：R1和R3上IPv6端不能同时抑制RA: ipv6 nd ra suppress
| interface Tunnel0 no ip address no ip redirects ipv6 enable tunnel source Ethernet0/1 tunnel mode ipv6ip 6to4 | interface Tunnel0 no ip address no ip redirects ipv6 enable tunnel source Ethernet0/2 tunnel mode ipv6ip 6to4 | 没有指定隧道的目的 根据目的IPv6地址推算出目的IPv4地址，然后和目的IPv4地址建立隧道。 |
特别情况：
两端IPv6网络不用2002:前缀。2002:前缀用到tunnel接口上
| interface Tunnel0 no ip address no ip redirects ipv6 enable tunnel source Ethernet0/1 tunnel mode ipv6ip 6to4 ipv add 2002:a0a:c01:1::1/64 | interface Tunnel0 no ip address no ip redirects ipv6 enable tunnel source Ethernet0/2 tunnel mode ipv6ip 6to4 ipv dd 2002:a0a:1703:1::1/64 | 没有指定隧道的目的 根据目的IPv6地址推算出目的IPv4地址，然后和目的IPv4地址建立隧道。 (将隧道接口当成内网IPv6接口，就是上面的案例了) |
| R1 | ipv route 2023::/16 2002:a0a:1703:1::1 ipv route 2002:a0a:1703::/48 tunnel0 | |
| R3 | ipv route 2022::/16 2002:a0a:c01:1::1 ipv route 2002:a0a:c01::/48 tunnel0 | |
4、自动隧道-----ISATAP隧道(略)
5、自动隧道-----IPv4兼容IPv6隧道(略)
6、半自动隧道-----6PE(后面实验部分介绍)
实验七、NAT--PT技术 (与CEF冲突，需关闭CEF)
A、静态NAT-PT(IPv4->IPv6单向)
interface Ethernet0/0
ipv6 address 2022::FFFF/64
ipv6 enable
ipv6 nat
interface Ethernet1/0
ip address 10.10.10.254 255.255.255.0
ipv6 enable
ipv6 nat
ipv6 nat v6v4 source 2022::10 10.10.10.1
ipv6 nat prefix 2023::/96
//
10.10.10.10->10.10.10.1 源地址10.10.10.10不会自动用前缀2023::/96构造新的IPv6地址作为源地址 2023::0a0a:0a0a去访问2022::10
必须添加映射：ipv6 nat v4v6 source 10.10.10.10 2023::1
B、静态NAT-PT(IPv4<->IPv6双向)
ipv6 nat v6v4 source 2022::10 10.10.10.1
ipv6 nat v4v6 source 10.10.10.10 2023::1
反向:
C、动态NAT-PT(IPv6->IPv4)
ipv6 access-list test
permit ipv6 2022::/64 any
ipv6 nat v6v4 pool pool1 10.10.10.200 10.10.10.253 prefix-length 24
ipv6 nat v6v4 source list test pool pool1
ipv6 nat v4v6 source 10.10.10.10 2023::1
D、动态NAT-PT(IPv4->IPv6)
access-list 1 permit 10.10.10.0 0.0.0.255
ipv6 nat v4v6 pool pool1 2023:: 2023::FFFF prefix-length 96
ipv6 nat v4v6 source list 1 pool pool1
ipv6 nat v6v4 source 2022::10 10.10.10.1
E、IPv4-Mapped NAT-PT(自动映射-IPv4映射到IPv6)------ IPv6->IPv4
手动将IPv4地址映射到IPv6，如果数量不多，还可以承受。如果有大量的IPv4地址映射到IPv6中，则可以使用自动映射。仅限IPv6主动访问IPv4。
ipv6 access-list test
permit ipv6 2022::/64 any
ipv6 nat v6v4 pool pool1 10.10.10.200 10.10.10.253 prefix-length 24
ipv6 nat v6v4 source list test pool pool1
ipv6 access-list v4map
permit ipv6 2022::/64 2023::/96
ipv6 nat prefix 2023::/96 v4-mapped v4map
当2022::/64 访问2023::/96时，目的ipv6地址直接去掉前面的96位前缀变成32位，将32位翻译成IPv4地址，，源地址使用 pool1中未使用的一个IPv4地址
2023::0a0a:0a0a去掉前面的96位，剩下0a0a:0a0a，翻译成IPv4地址就是10.10.10.10
实验八、6PE技术
前面介绍的隧道技术，其实是将IPv6包封装在其他协议中(GRE或者IPv4)，然后通过承载网路由到目的网络。
6PE技术，在两个PE之间建立IPv4 BGP对等体，但是传递的是IPv6路由。
IPv6数据包在承载网中传输时，会在包前打上两层标签。
| PE1: router bgp 234 bgp router-id 2.2.2.2 bgp log-neighbor-changes no bgp default ipv4-unicast neighbor 4.4.4.4 remote-as 234 neighbor 4.4.4.4 update-source Loopback0 ! address-family ipv4 exit-address-family ! address-family ipv6 redistribute connected redistribute ospf 1 neighbor 4.4.4.4 activate neighbor 4.4.4.4 send-label exit-address-family | PE2: router bgp 234 bgp router-id 4.4.4.4 bgp log-neighbor-changes no bgp default ipv4-unicast neighbor 2.2.2.2 remote-as 234 neighbor 2.2.2.2 update-source Loopback0 ! address-family ipv4 exit-address-family ! address-family ipv6 redistribute connected redistribute ospf 1 neighbor 2.2.2.2 activate neighbor 2.2.2.2 send-label exit-address-family |
CE1和CE2都学到了对方的路由
IPv6包前面加了两层标签，外层标签300由P设备产生，通过LDP协议传递过来；内层标签403由PE2的MP-BGP产生，通过MP-BGP传递过来(默认是不产生标签的，需要通过命令 neighbor 4.4.4.4 send-label 和neighbor 2.2.2.2 send-label来产生标签).
为什么要两层标签？
MPLS标签协议中有一个倒数第二跳标签弹出机制。如果只有一层标签，那么在P设备（倒数第二跳）上弹出标签后，只剩下IPv6数据包了，P设备上没有启用IPv6，不可能路由，只能丢弃数据包了。
实验九、6VPE技术
6PE技术，在两个PE之间建立IPv4 BGP对等体，但是传递的是IPv6路由。
6VPE技术，在两个PE之间建立VPNv4 BGP对等体，但是传递的是VPNv6路由。
| PE1 | PE2 |
| vrf definition cisco rd 1:1 ! address-family ipv6 route-target export 234:2 route-target import 234:4 exit-address-family | vrf definition cisco rd 2:2 ! address-family ipv6 route-target export 234:4 route-target import 234:2 exit-address-family |
| interface Ethernet0/0 vrf forwarding cisco no ip address duplex auto ipv6 address 2001:12::2/64 ipv6 enable | interface Ethernet0/0 vrf forwarding cisco no ip address duplex auto ipv6 address 2001:45::4/64 ipv6 enable |
| router bgp 234 bgp router-id 2.2.2.2 bgp log-neighbor-changes no bgp default ipv4-unicast neighbor 4.4.4.4 remote-as 234 neighbor 4.4.4.4 update-source Loopback0 ! address-family ipv4 exit-address-family ! address-family vpnv6 neighbor 4.4.4.4 activate neighbor 4.4.4.4 send-community extended exit-address-family ! address-family ipv6 vrf cisco redistribute connected neighbor 2001:12::1 remote-as 100 neighbor 2001:12::1 activate neighbor 2001:12::1 as-override exit-address-family | router bgp 234 bgp router-id 4.4.4.4 bgp log-neighbor-changes no bgp default ipv4-unicast neighbor 2.2.2.2 remote-as 234 neighbor 2.2.2.2 update-source Loopback0 ! address-family ipv4 exit-address-family ! address-family vpnv6 neighbor 2.2.2.2 activate neighbor 2.2.2.2 send-community extended exit-address-family ! address-family ipv6 vrf cisco redistribute connected neighbor 2001:45::5 remote-as 100 neighbor 2001:45::5 activate neighbor 2001:45::5 as-override exit-address-family |
PE和CE之间使用BGP协议，因为两端都是AS100，因此需要as-override
CE1和CE2都能学到路由。
IPv6包前面加了两层标签，外层标签301由P设备产生，通过LDP协议传递过来；内层标签403由PE2的MP-BGP自动产生，通过MP-BGP传递过来.
注：有几张图片是用的别人的。
1万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
